Large Type Fit Indices of Mathematics Adult Learners: A Covariance Structure Model

Abstract

Fit is the ability of a model to reproduce the data in the variance-covariance matrix form. A good fitting model is one that is reasonably consistent with the data and doesn’t require respecification and also its measurement model is required before estimating paths in a covariance structure model. A baseline model of four constructs together with a combination of none, one, two, three or four additional constructs was constructed with latent variables: educational performance, socio-economic label, self concept and parental authority using dichotomous digits 0 or 1 for each additional construct. We considered 16 progressively nested models starting with baseline model using the mathematics adult learners data from the modeling sample and employing some large fit indexes which are commonly used (NFI, NNFI, CFI, GFI, PGFI, among others) Usluel, et al. (2008) to test the fitness of the model. The measures of model fit based on results from analysis of the covariance structure model are presented

Keywords: Fit Indices; Structural Equation Modeling; Bernoulli Digits; Latent Constructs; Educational Performance